comparative notes on the madhyama-āgama

8/10/2019 Comparative Notes on the Madhyama-gama

1/24

Comparative Notes on the Madhyama-gamaAnlayo

Abstract:The present paper offers a survey of some features of the Madhyama-gama,based on a comparison with its extant parallels. After taking up matters related tothe structure of the Madhyama-gama collection, selected passages from

Madhyama-gama discourses will be examined in order to show the importanceof placing the different versions of a discourse, preserved in Chinese and Pli (and

at times also in Sanskrit or Tibetan) side by side, in order to ascertain theimplications of certain passages and to avoid being misled by transmission ortranslation errors.

According to the information that has come down to us, theMadhyama-gamacollection was translated towards the end of the fourth century into Chinese underthe leadership of the Kashmirian monk Gautama Saghadeva. The translation wasbased on a written Indic original read aloud by Sagharaka, another Kashmirianmonk, and was transcribed by the monk Do-c, with the assistance of L-bo and Kng-hu, and .1The Indian original used for this translation

appears to have been in a Prkrit,2

and with considerable probability stems from aSarvstivdatradition.3

The Madhyama-gamacollection contains altogether two-hundred-and-twenty-two discourse, which are assigned to eighteen chapters. Each of these chapterscontains a minimum of ten discourses, though a few chapters have considerablymore. Its Pli counterpart, the Majjhima-nikya, contains one-hundred-fifty-twodiscourses in fifteen chapters, arranged into fourteen chapters with ten discourseseach and one chapter with twelve discourses. Hence, in spite of some variationsthe principle of ten discourses per chapter could be an original characteristic ofboth collections, a grouping principle that is in fact recurrent in Buddhist literature.

1T I 809b26: (with a variant reading for the last as ).

2On the language of theMadhyama-gamamanuscript cf. Bapat 1969: 5; Enomoto 1986: 20 andvon Hinber 1982: 250.3On the school affiliation of theMadhyama-gamacf. Mayeda 1985: 98 and Minh Chau 1991: 27.Enomoto 1984: 198 explains that the Madhyama-gama translated into Chinese probablyrepresents the earliest of three versions of this collection, the second of the three being theversion preserved in some of the Central Asian Sanskrit fragments and the third version beingwhat is found in straquotations in later works.

33


2/24

Regarding the chapter division in the two collections, four chapters in the Madh-

yama-gama and the Majjhima-nikya share the same headings and also haveseveral discourses in common. These are the chapters on kings, on Brahmins, onexpositions (vibhaga), and on pairs; chapters that occur, however, at differentplaces in the two collections.4Two Chinese discourse from the chapter on kings;four Chinese discourses from the chapter on Brahmins and from the chapter onpairs; and nine Chinese discourses from the chapter on expositions have a parallelin their Pli equivalent chapter.

Figure 1:Discourse parallels in similarly entitled Madhyama-gama and Majjhima-nikyachapters

Chapters:

Discourses: M 63 / MN 81M 67 / MN 83

M 150 / MN 96M 151 / MN 93M 152 / MN 99M 161 / MN 91

M 162 / MN 140M 163 / MN 137M 164 / MN 138M 165 / MN 133M 166 / MN 134M 167 / MN 132M 169 / MN 139M 170 / MN 135M 171 / MN 136

M 182 / MN 39M 183 / MN 40M 184 / MN 32M 185 / MN 31

During the progress of my research, it has been possible to update theidentification of parallels between Madhyama-gama and Majjhima-nikyadiscourses noted in Akanuma 1990. I differ from Akanuma in regard to M 28;M 86; M 106 and M 168. Akanuma 1990: 171 lists M 28 as a parallel toMN 143. But M 28 agrees with SN 55.26 at SN V 380 as regards Sriputtasinstructions to Anthapiika, and in having Anthapiika recover, while in MN143 he passes away and the instructions he receives are also different. Hence, M28 is better reckoned a parallel to SN 55.26. Akanuma 1990: 171 lists M 86 as aparallel to MN 148. In M 86, nanda asks the Buddha how to instruct a groupof young monks. The Buddha responds by discussing the five aggregates, the sixsenses, dependent origination, the four establishings of mindfulness, the four right

efforts, the four ways to power and a whole range of other topics. In contrast, in

4The 6thchapter in theMadhyama-gama, the chapter on sayings "connected with kings" (), has a counterpart in the 9thchapter in theMajjhima-nikya, theRja-vagga. The 12thchapterin the Madhyama-gama, the chapter on "Brahmins" ( ), has its counterpart in theBrhmaa-vagga, the 10th chapter in the Majjhima-nikya. The 13thchapter in the Madhyama-gama, the chapter on "expositions" () has its counterpart in the Vibhaga-vagga, the14thchapter in the Majjhima-nikya. The 15thchapter in the Madhyama-gama, the chapter on"pairs" (), has a counterpart in the Mahyamaka-vagga, the 4th chapter in the Majjhima-nikya. Cf. also Yn-shn 1983: 707.

34


3/24

MN 148 nanda does not occur at all and the topic of the discourse are only the

six senses. Thus, M 86 and MN 148 differ to such a degree that they cannot bereckoned parallels. Akanuma 1990: 169 lists M 168 as a parallel to MN 120. Yet,closer examination shows that the two discourses differ considerably from eachother, as MN 120 describes how a mental aspiration can lead to various rebirths,while M 168 describes how jhna practice leads to the Brahm worlds. Thismakes it improbable that the two discourses stem from the same original and canbe reckoned as proper 'parallels', as in spite of similarity of topic they seem to goback to different occasions. The same applies also to M 106 (and T 56), whichAkanuma 1990: 163 lists as parallels to MN 1. Closer inspection shows thatwhereas MN 1 discusses worldling, disciple in higher training, arahant, andtathgata, M 106 and T 56 discuss two types of recluses/Brahmins and the

Buddha. They name a different location for the discourse and differ from MN 1 innot discussingNibbna, and in not concluding with the monks failing to delight inthe exposition. In contrast, E 44.6 agrees with MN 1 on the above points.Therefore, M 106 and T 56 also do no seem to qualify for being parallels in theproper sense. Thus, I would conclude M 28 and M 86 are definitely notparallels toMajjhima-nikya discourses; and in the case of M 106 and M 168 Ihave strong doubts if these should be reckoned as 'parallels'. In addition to theparallels recognized by Akanuma, M 29 is a parallel to MN 9,5and M 208 is aparallel to MN 79.

Thus, according to my reckoning ninety-five Majjhima-nikya discourses havecounterparts in the Madhyama-gama, which, as two Madhyama-gamadiscourses are counterparts to a single Majjhima-nikya discourse,6count up toninety-six discourses. These parallels are arranged in the two collections in ratherdifferent ways. This difference in arrangement can best be illustrated by placingthe Majjhima-nikya discourses in the order their parallels occupy in the

Madhyama-gama, as done in figure 2. Even a cursory glance at figure 2 showsthe extent to which the arrangement of discourses differs in the two collections.These differences support the impression that the location of the discourses wasthe outcome of a process specific to each of the two collections, though thesimilarities exhibited in figure 1 could be the remnants of a common starting-point.7

5See Yin-shun 1983: 709.6M 107 and M 108 are both parallels to MN 17.7In a similar vein, in regard to the Sayutta-nikya and Sayukta-gama collections Glass 2006:6 comes to the conclusion that while the "shared principle of arrangement is likely to be very old,important differences between the content and arrangement of the extant versions show that theyfollowed separate developments". Norman 1992: 40 explains that "the sects ... had the samenames for the groups of texts, but were not ... in general agreement about their contents, or theorder of the contents".

35


4/24

Figure 2:

Majjhima-nikya discourses arranged in the sequence of their Madhyama-gamaparallels

M chapters: MN discourses:1

stchapter - - - - - - - - 24

th2

nd

2nd chapter - - - 61st - - - - 101st -

3rd chapter - - - - - 69th 97th - 9th 28th 141st

4th chapter 123rd - 124th - - - - - - -

5th chapter - - - - - - - - - - - - - - - -

6th chapter - - - - - 81st 130th - - 83rd - - - -

7th

chapter 128th

- - 106th

- 68th

49th

127th

- 119th

- - - 113th

-

8th

chapter 5th

3rd

15th

- 8th

- 7th

- - -

9th

chapter - 10th

13th

14th

20th

19th

11th

- 6th

-10th chapter 17th 17th - - - - - - 18th -

11th

chapter - - - - - - - - - - - - -

11th

chapter - 50th

82nd

56th

- - - - - - - -

12th

chapter - - 107th

108th

27th

- - - 96th

93rd

12th

chapter 99th

75th

- - - - - - - 91st

13th chapter 140th 137th 138th 133rd 134th 132nd - 139th 135th 136th

14th

chapter - 126th

45th

46th

- - 25th

78th

142nd

115th

15th

chapter 39th

40th

32nd

31st

47th

112th

- 117th

121st

122nd

16th chapter 66th 21st 65th 70th 104th - 125th 129th 22nd 38th

17th

chapter - 54th

26th

64th

16th

77th

79th

80th

44th

43rd

18th

chapter 90th

89th

88th

- 87th

52nd

- - - 63rd

-

A structural similarity can be found in the cross-tradition relationship betweenthe middle collections and the numerical collections. The percentage of

Aguttara-nikya parallels toMadhyama-gama discourses is almost the same asthe percentage ofEkottarika-gama parallels toMajjhima-nikya discourses.8

Another pattern of similarity can be found in relation to the (Mla-)SarvstivdaDrgha-gama collection, preserved in Sanskrit fragments, as the number ofparallels to Majjhima-nikya discourses found in thisDrgha-gama collectioncorresponds to the number of parallels to Dgha-nikya discourses found in the

Madhyama-gama, both being ten, as can be seen in figure 3 below.9In spite ofthis parallelism, a closer look at figure 3 shows that the actual order in whichthese parallels occur in both cases does not seem to follow a specific pattern, nor

do these two sets of ten parallels appear to be related to each other.

8 A survey of the parallels noted by Akanuma 1990: 7-25 suggests that about 37% of thediscourses in the Madhyama-gama collection have a parallel in the Aguttara-nikya; whileabout 35% of the dis courses found in theMajjhima-nikya have a partial or a full parallel in theEkottarika-gama (the second figure is based on my own research into Majjhima-nikyaparallels).

9Of these ten Drgha-gama discourses, eight do not appear to have a parallel in any of the fourChinese gamas. Only D211 has angama parallel, which is E 31.1; while D212 has aparallel in an individual translation outside of thegamas, T 757.

36


5/24

Figure 3:10

Parallels to DNdiscourses found in M

and parallels to D2

discourses found in MN

To sum up, the patterns of similarity that can be discerned are:

1) The percentage of Madhyama-gama parallels in the Aguttara-nikyamir-rors the percentage ofMajjhima-nikya parallels found in theEkottarika-gama.

M AN MN E2) The number of Madhyama-gama parallels in the Dgha-nikya is the same asthe number of Majjhima-nikya parallels in the Sanskrit Drgha-gama collec-tion.

M DN =MN D2

While it cannot be excluded that such similarities are a product of chance, theyare striking enough and one would not expect to come across such similaritiesunless the basic four-fold division into four Nikyas or gamas stems from acommon source.12

The supposition that theMadhyama-gama and theMajjhima-nikya stem froma common starting point is also a prominent impression to be gained whenindividual discourses in these two collections are compared to each other. It is

10The second part of this table has been adopted from Hartmann 2004: 126-127.11Hartmann 2000: 365 note 20 indicates that this identification still needs to be corroborated.12

While Bechert 1991: 9 believes that "the compilations available to us hardly go back to any 'Ur-gamas', but originated as the result of local applications of the same principles of organisation",suggesting that the similarities among the extant collections then "lead to the erroneous assump-tion that there might have been an original form of the corpus as a whole"; Norman 1989: 33suggests that "probably ... from the earliest period of Buddhism the collection of sermons wasmade on the basis" of these four basic groups, which "was probably the beginning of the systemof bhakas ('reciters'), who shared out the recitation of the various sections of the Buddha'steaching among themselves". Prasad 1985: 137 comments on the arrangement of discourses inthe canons of various schools that "the criteria for such distribution were clear but the discoursesare by nature such that they satisfy more than a single criteria. In [the] course of time theposition of those discourses ... changed in different schools".

Dgha-nikya Madhyama-gama Skt.Drgha-gama Majjhima-nikyaDN 15 M 97 D27 MN 6011DN 17 M 68 D210 MN 105DN 21 M 134 D211 MN 4DN 22 M 98 D212 MN 12DN 23 M 71 D217 MN 102DN 25 M 104 D219 MN 95DN 26 M 70 D220 MN 36DN 27 M 154 D221 MN 85DN 30 M 59 D222 MN 100DN 31 M 135 D243 MN 55

37


6/24

quite amazing to find that the version of a discourse recited and written down

perhaps four centuries after the Buddha's passing away by Sri Lankan monks canbe so closely similar,13even in small circumstantial details, to a discourse handeddown by a different Buddhist school and translated another four centuries laterinto Chinese. At times, such similarities even involve sharing the same mistakes.Yet, at the same time there are undeniably quite a number of differences and

variations, as is only to be expected of orally transmitted material.Another striking impression that can be gained from a comparative study of the

Madhyama-gama is the relatively high quality of its translation. Though thistranslation still falls into the early phase of translation activity in China, onreading this collection one gets the strong impression that the team aroundSaghadeva knew what they were doing and endeavoured to render the Indic text

at their disposal to the best of their abilities. Thus, the Madhyama-gamacollection is certainly a good instance to corroborate the statement madeLancaster (1979: 224) that "in the Chinese canon we have an invaluable source ofevidence ... with some assurance that those translators knew their craft andpracticed it with vigour and accuracy". Thus, as de Jong (1968: 15) points out, "nostudent of Buddhism, even if he is interested only in Indian Buddhism, canneglect the enormous corpus of Chinese translations".Nevertheless, translation errors inevitably occurred, and a description of what

appear to be such errors, together with what may be transmission errors thataffected either the Chinese version or its Pli parallel, will be the theme of theremainder of my presentation.

The first example I would like to take up stems from the , the 14thdiscourse in theMadhyama-gama. This discourse records an instruction given bythe Buddha to his son. As part of this instruction, the Buddha tells his son that heshould refrain from a bodily action that is "pure" 14and at the same time unwhole-some and results in affliction; whereas a bodily action that is "not pure" but at thesame time is wholesome and does not result in suffering can be undertaken.

1)1516

This presentation is surprising, since to qualify an action as wholesome and atthe same time as impure seems contradictory. Minh Chau (1991: 34 and 132)

13Regarding the date of the Buddha's passing away, the research collected in Bechert 1995suggests a reasonable date to be approximately 400 B.C. (allowing a latitude of +/- twentyyears).

14Lvi 1896: 480 renders in the present context as "pur".15M 14 at T I 436c11.16M 14 at T I 436c14.

38


7/24

suggests that the character in this context could refer to actions that are "per-

missible" according to the monastic code of discipline but have unwholesome re-sults, as opposed to actions that are not permissible but that have wholesomeresults. He then concludes that this instruction allows a more liberal attitudetowards the interpretation of monastic regulations. Minh Chau's reasonableattempt to make sense out of this passage by assuming that may have ameaning different from its more usual meaning of "purity" receives support fromthe Chinese-Sanskrit dictionary compiled by Hirakawa (1997: 728), who inaddition to uddha, uddhi, pariuddhi, viuddhi, pariodhayati, ubha,vyavadna, prasanna, vimala also lists kalpika, "proper", and kalpa, "proper,practicable, feasible, possible", as equivalents for . A problem with MinhChau's interpretation, however, is that according to the instruction for a pastbodily action in the a "permissible" () bodily deed should be confessed,while a "not permissible" () bodily deed leads to the arising of joy.17Thisstatement makes no sense, since for a "permissible" bodily deed there would be noneed for confession. In fact, the idea to give precedence to other's welfare over therequirements of moral conduct belongs to a later phase of Buddhist thought anddoes not yet seem to be found in the early discourses.

The of the Madhyama-gama has a parallel in the Ambalahikrhu-lovda-sutta of the Majjhima-nikya. The corresponding passage in the Pli

version reads differently, since it does not envisage that a bodily action could bepure and unwholesome at the same time, but simply classifies such an

unwholesome bodily deed as a deed that results in affliction.18

The same is alsothe case for another parallel to the found in the (Mla-)SarvstivadaVinayavibhaga, preserved in Chinese and Tibetan.19A stra quotation from thepresent discourse in the Vykhyyukti-k, preserved in Tibetan, 20 and aparaphrase of the present passage in the rvakabhmi, agree with the Pli dis-course and the Vinayavibhagaversion that the present instructions are to refrainfrom a bodily deed that is harmful and unwholesome,21without envisaging thatsuch a deed could be reckoned as pure.

In fact, if the instruction to Rhulahad offered such an ambivalent instruction,one would not expect it to feature among King Aoka'sexplicit recommendations,a choice that might well be due to the straightforward and practical ethical instruc-

17M 14 at T I 436c27.18MN 61 at MN I 415,29: akusala ida kyakamma dukkhudraya dukkhavipka.19T 1442 at T XXIII 761a11: . D 'dul ba cha217a56or Qje

201a6: bdag dang gzhan la gnod par 'gyur ba mi dge ba(D: ba'i) sdug bsngal 'byung ba rnampar smin pa sdug bsngal ba.

20Q sems tsam i 71a5:gnod pa dang ldan pa mi dge ba sdug bsngal 'byung ba dang.21T 1579 at T XXX 405b5: , or Shukla 1973: 55,16: vybhdhika ... tmano

v parasya v akuala.

39


8/24

tion given in this discourse.22 A king would quite probably not recommend a

discourse that enjoins doing what is not permitted because one believes it to bewholesome. Thus, it seems that this part of the Madhyama-gama version mighthave suffered from a translation error. Such an error could have occurred due tomisinterpreting a sandhiin the Indic original to imply that a particular word has,or else does not have, the negative prefix a-. A similar error can be found in the194thdiscourse of theMadhyama-gama,23and Karashima (1992: 263) notes theoccurrence of this type of error in Dharmaraka's translation of the Saddharma-

puarka-stra, confirming that such mistakes did take place.

The next example is related to the translation terminology employed in theMadhyama-gama, taken from the 25th discourse, the . This discourse

refers to a forest dweller, araaka/arayaka, in a way that seems to be quiteunique in the Madhyama-gama collection, namely as, 24instead of themore usual rendering of the same as "no thing", .25This variation from theusual Madhyama-gama translation vocabulary is interesting in so far as thetranscription occurs with considerable frequency in the Ekottarika-gama.26 This detail thus could be of relevance to the relationship between thetranslations of these two discourse collections. The identity of the translator of the

Ekottarika-gamais not an unequivocal matter, since it is not entirely clear if thetranslation now extant in Chinese has only been revised by Gautama Saghadeva,or whether it is an actual retranslation undertaken by him, a retranslation that then

replaced Dharmanand's earlier translation.

27

A problem with this suggestion is that in general the translation terminology inthese two collections differs considerably, and that to such an extent that makes itdifficult to assume how the two collections could have been translated under thesame person.

It is also not clear on what Gautama Saghadeva would have based such a

22This recommendation refers to the Lghulovda, "spoken by the Blessed One, the Buddha,concerning falsehood", mus vda adhigichya bhagavat budhena bhsite, cf. Hultzsch 1925:173.

23Cf. below quote number 13 (M 194).24

M 25 at T I 454a19.25Noted by Minh Chau 1991: 327 as the standardMadhyama-gamarendering for araa.26Cf. e.g. E 12.5 at T II 569c14; E 12.6 at T II 570a25; E 13.1 at T II 571b2; E 25.6 at T II

633b16; E 37.3at T II 711a8; E 38.6 at T II 721a1; E 39.10 at T II 734a9; E 49.2 at T II795a26.

27The , T 2145 at T LV 71b29; the , T 2146at T LV 127c29; the , T 2153 at T LV 422b6; and the , T 2154 at T LV 511b15, attribute theEkottarika-gamatranslation to Dharmanand; while according to the , T 2034 at TXLIX 70c5, Gautama Saghadeva retranslated the Ekottarika-gama; cf. also Anlayo 2006;Bagchi 1927: 159 and 337; Enomoto 1986: 19; Lamotte 1967: 105; L 1963: 242; Mayeda 1985:102; Waldschmidt 1980: 169 note 168; and Yin-shun 1983: 93.

40


9/24

retranslation, since whereas in the case of the Madhyama-gamahis translation

was based on a written original, Dharmanand translated the Ekottarika-gamabased on an original he had memorized,28and there is no indication that GautamaSaghadeva had also memorized this collection or had otherwise access to anoriginal corresponding to what Dharmanand had committed to memory.Nevertheless, the occurrence of the term in theMadhyama-gama points

to some form of relationship between the translation of these two collections, infact we know that Dharmanand had earlier also translated a Madhyama-gamainto Chinese, though it seems that this translation was subsequently lost. Thepresent finding thus leaves open the possibility, suggested by L (1963: 242), thatsome of the renderings employed by Dharmanand in his earlier Madhyama-gama translation were re-employed in Saghadeva's Madhyama-gama

translation.

The next example I would like to take up appears to be a simple translation error,found in the 26th discourse in the Madhyama-gama, the . In adescription of proper behaviour for monks in regard to seats, this discourseinstructs that one should not encroach on elder monks and younger monks shouldnot be "scolded", .

2)29

That when taking a seat one should not incommode elder monks is certainlyreasonable, but why there should be any scolding of younger monks is less easy tounderstand. According to the Pli parallel, the Gulissni-sutta, the instruction inthe present case it to not "keep off" young monks from their seats, nave cabhikkh na sanena paibhissmi.30This suggests the possibility of a translationerror, which could have happened due to mistaking paibhati/pratibdhate, "tokeep off", forpaibhsati/pratibhate, "to retaliate".

The next discourse in the Madhyama-gama, the , describes a visitpaid by Sriputta to a Brahmin. According to the Madhyama-gama discourse,when Sriputta arrived he found that this Brahmin was outside of his house by the

side of a spring and was "inf licting pain on the resident people".3)31

In this case, too, a translation error appears to have happened. According to the

28T 2145 at T LV 10b25.29M 26 at T I 455c6.30MN 69 at MN I 469,18.31M 27 at T I 456c9.

41


10/24

Pli parallel, the Dhnajni-sutta, when Sriputta arrived this Brahmin was at

his cowshed getting his cows milked, gvo gohe dohpeti.32

The idea of"inflicting pain", , could to be due to a mistaking of duh, "to milk", for dru,"to harm" or "to hurt". The occurrence of "resident people", , is less easilyexplainable, though perhaps the idea that someone is being harmed or hurt mighthave led to a misinterpretation of goha/goha, "cow-pen", for gotta/gotra,"clan". Alternatively, the reference to the "resident people" could be a glossintroduced by the translator.

A mistaking of duhfor dru could occur more easily in a Prkrit in which, likein Pli, the two forms are not distinguishable by the occurrence of an r in the latter,whereas such a mistaking would seem less probable in Sanskrit. Thus, thistranslation error would support the assumption that the original used fortranslating theMadhyama-gama was in a Prkrit.

These few examples already show the importance of studying the earlydiscourses in conjunction, in the sense of reading the different versions availablealongside each other. This is certainly the case in both ways, that is, not only doesa reading of a Madhyama-gama discourse benefit from examining its Pliparallel, but similarly a reading of a Plidiscourse benefits from an examinationof its Chinese counterpart.33This is because errors are not only to due translation,but can also be the outcome of lapses of memory during oral transmission. Anexample for this potential can be found in the 63rd discourse in the Madhyama-gama, the . This discourse describes a situation where the formerBuddha Kassapa sent some of his monks to the house of one of his supporters inorder to procure grass for thatching his roof. When the monks arrive, thesupporter himself was out and only his blind parents were at home. The

Madhyama-gamaversion reports that the blind parents asked who had come, andthe monks replied by explaining who they are.

4) 34

The Pli version of this exchange differs in so far as it reports that the monks ad-dressed the blind parents with the word "sister", bhagini.35 According to the

Madhyama-gamaversion, however, they instead used the address ,corresponding to yasmant/yumant, a respectful form of address regularly usedto refer to monks, but also appropriate towards elders in general. In a patriarchalsociety like ancient India, where the wife played a subordinate role comparable to

32MN 97 at MN II 186,2(Be-MN II 396 and Se-MN II 624 read dhpeti).33A survey of examples for this potential of the Chinesegamascan be found in Anlayo 2005.34 M63 at T I 502b24.35MN 81 at MN II 53,25.

42


11/24

a servant,36one would not expect that monks who speak to a couple would only

address the woman, so that the form of address given in theMadhyama-gamaisclearly the preferable reading.

My next example is from the 78thdiscourse in the Madhyama-gama, the , and its parallel in theBrahmanimantaika-sutta. This discourse describesan encounter between the Buddha and a conceited Brahm, who mistakenlythought himself to be eternal. In order to dispel this deluded assumption ofBrahm, the Buddha pointed out that he knew where Brahm had come from andwhere Brahm was going to, thereby showing that Brahm's present existence wasfar from being eternal.

5)37

In the Brahmanimantaika-sutta, however, the Buddha told Brahm that heknew Brahm's destiny and "splendour", te aha, Brahme, gatica pajnmi

jutica pajnmi.38The occurrence ofjuti, "splendour", is puzzling, since thoughthe Buddha would know all about Brahm, and therewith also all that is to beknown about Brahm's splendour, to display such knowledge would not be aseffective a challenge to Brahm's belief in being eternal as theMadhyama-gama

version's proposal that the Buddha knew where Brahm was going to. Knowledgeof where Brahm was going to would indeed undermine Brahm's belief in beingeternal, whereas knowledge of Brahm's splendour would not seem to be sopertinent to the point at stake in the present context.

TheMadhyama-gama version receives support from a variant reading found inthe PTS and the Sinhalese edition, which reads cuti, "passing away".39In view ofthe context this appears to be the preferable reading. In this way, the Madhyama-gamaversion helps us to decide in favour of a reading found only as a variant inthe Pli editions.

An aspect of the early discourses that shows considerable variations betweendifferent versions of the same discourse is the title. This is not only the case for

versions from different reciter traditions, since even between Pli editions of the

36Von Hinber 1993: 102 draws attention to Vin IV 21,3, where a mother-in-law addresses herdaughter-in-law withje, an address elsewhere used to address a female slave, cf. e.g. MN 21 atMN I 125,18. This form of address reflects the low social position of a daughter-in-law in thehousehold of her in-laws. Horner 1990: 1 explains that in ancient India a married women's "lifewas spent in complete subservience to her husband and his parents. She was allowed littleauthority at home and no part in public activities."

37M 78 at T I 548a8.38MN 49 at MN I 328,25.39PTS edition at MN I 557 and Ce-MN I 768 note 5; cf. alsoHorner 1967: 391 note 5, who com-

ments thatjuti "seems to be faulty for cuti".

43


12/24

same discourse a considerable degree of variation can be found in regard to titles.

In theMajjhima-nikya collection, for example, more than ten discourses have asubstantially different title in another Pli edition.40

In view of such variations, it is not surprising that at times the title of a Madh-yama-gama discourse appears to be preferable to the title of its Majjhima-nikyaparallel. Such a case can be found in the 101stdiscourse in theMadhyama-gama,the . 41 Its counterpart is the Vitakkasahna-sutta in the Majjhima-nikya. Now vitakkasankhrasahna, "stilling the thought-formation",42is onlyone of the five methods described in this discourse for overcoming unwholesomethoughts, whereas all five methods are for the purpose of developing the "highermind", adhicitta, .43Thus in as much as a title for the whole discourse isconcerned, the Madhyama-gama version's title seems to fit better than its Plicounterpart.

Another aspect of the early discourses that shows considerable variationsbetween different versions is the sequence in which otherwise similar aspects orteachings are presented. A case in point is the 102nddiscourse of theMadhyama-gama, the , which describes how the Buddha, during the time before hisawakening, handled the arising of unwholesome thoughts. This discoursedescribes that whenever an unwholesome thought arose, the future Buddha wouldquickly dispel such thoughts, as he was aware of the danger inherent in them. The

Madhyama-gama discourse compares this to a cowherd who would stop thecows from straying into the ripe crop, as he knows that he will incur trouble if hedoes not prevent them from eating the crop. 44 After explaining this method,according to the Madhyama-gama discourse the Buddha turned to the generalnature of the mind, explaining that whatever one frequently thinks about willeventually lead to a corresponding inclination of the mind.45

Its Pli counterpart, theDvedhvitakka-sutta, however, presents these topics in a

40 These are the Vatthpama-sutta, MN 7, where Be has the title Vattha-sutta; the Ariyapari-yesana-sutta, MN 26, where Be and Sehave the title Psarsi-sutta; the Sekha-sutta, MN 53,where Se has the title Sekhapaipad-sutta; the Upli-sutta, MN 56, where Se has the titleUplivda-sutta; the Ambalahikrhulovda-sutta, MN 61, where Se has the titleCarhulovda-sutta; the Camlukya-sutta, MN 63, where Se has the title

Camlukyovda-sutta; the Tevijjavacchagotta-sutta, MN 71, where Be

has the title Tevijja-vaccha-suttaand Sethe title Cavacchagotta-sutta; theAggivacchagotta-sutta, MN 72, whereBe has the title Aggivaccha-sutta; the Mahvacchagotta-sutta, MN 73, where Behas the titleMahvaccha-sutta; the Bakkula-sutta, MN 124, where Se has the title Bakkulattherac-chariyabbhta-sutta; and the Mahsayatanika-sutta, MN 149, where Se has the titleSayatanavibhaga-sutta.

41M 101 at T I 588a3.42Adopting the rendering used in amoli 2005: 212.43MN 20 at MN 119,3introduces the five methods with adhicittam anuyuttena ... bhikkhun.44M 102 at T I 589a25.45M 102 at T I 589b5.

44


13/24

different sequence, as it first takes up the dispelling of unwholesome thought, then

describes how the mind follows the course set by whatever one frequently thinksabout, and only after that comes out with the simile of the cowherd.46Here the

Madhyama-gama discourse present a more straightforward sequence, since thepurpose of the cowherd's simile is to illustrate fear of unwanted consequences, notto illustrate that frequent thoughts lead to a mental inclination. Thus, the simile ofthe cowherd finds its best placing right after the exposition of unwholesomethoughts, as an illustration of this exposition.

My next example is from the 145thdiscourse of theMadhyama-gama, the ,records a remark by nanda, in which he pointed out that the delightedin living in the Bamboo Grove because of the protection given by the Blessed One.

Since according to the same discourse the Buddha had already passed away by thetime nanda made this remark, one might wonder what the implications of thisprotection could be.

6)4748

According to the Pli version, however, nanda remarked that the conduciveconditions in the Bamboo Grove were due to the protection given by Vassakra, aminister of King Ajtasattu.49The Pli commentary then explains that Vassakragave particular care to the Bamboo Grove, as according to a prediction he wasgoing to be reborn in his next life as a monkey in this grove, a rebirth prospectivealso referred to in the Karmavibhaga.50 Thus, in this case it seems as if thereference in theMadhyama-gama version to the protection given by the BlessedOne, who had already passed away, may just be a simple transmission ortranslation error, and the protection spoken of in the present context appears tointend merely the mundane type of protection a minister might give to a park.

Another passage of interest to the present topic can be found in the 162nddiscourse in theMadhyama-gama, the ,lists and examines five typesof feeling tones that can be experienced, which are happiness, pain, mental

pleasure, mental displeasure and equanimity, but then summarizes these as threetypes of feelings.

46MN 19 at MN I 115,29.47M 145 at T I 653c22.48M 145 at T I 655b14.49MN 108 at MN III 13,20.50Ps IV 73 and Kudo 2004: 72,5and 73,4.

45


14/24

7) 51

Its Pli parallel, the Dhtuvibhaga-sutta, however, lists only three types offeelings in the corresponding passage, sukha, dukhhaand adukkhamasukha.52Inthe present case this reading is confirmed by the Madhyama-gama discourseitself, as in its summary of the present exposition it speaks of three types offeelings, so that the five-fold presentation appears to be a later expansion of whatoriginally was only a treatment of three types of feelings. Further confirmationcan be gleaned from another version of the same discourse, preserved as anindividual translation by Zh-qan, , which also refers only to three feelings. 53The present instance is interesting in so far as it documents the beginning stages inthe development of a tendency towards ever more detailed analysis, a tendency

that eventually led to the rise of the differentAbhidharmas.

The next discourse in the Madhyama-gama collection, the 163rd discourseentitled , takes up a related topic, as it examines the six types ofpleasure, of displeasure and of equanimity, presenting each set of six as eitherrelated to sensuality or else not related to sensuality. The resultant thirty-six foldpresentation receives the rather puzzling heading of being "thirty-six knives".

8) 54

The corresponding Pli passage, however, speaks of "thirty-six positions ofbeings", chattisa sattapad. 55 This suggests that the expression "thirty-sixknives", , could be due to mistaking satta/sattva, "being", forsattha/astra, "knife", or could perhaps be a copyist's mistake, confusing with . 56 The corresponding expression in the Abhidharmakoabhya readsatriac chstpadni, "thirty-six positions [set forth] by the teacher",57whichthus has satthar/st; "teacher", instead of satta/sattva, "being". This expressionrecurs in Xun-zng's () translation as ;58while Paramrtha's

51M 162 at T I 691b5-c5.52

M 140 at M III 242,11.53T 511 at T XIV 780b29.54M 163 at T I 692c16.55MN 137 at MN III 217,8.56The same expression recurs in Saghabhti's Vibh translation in T 1547 at T XXVIII 435c29:

, with as a variant for .57Abhidh-k 3:36 in Pradhan 1967: 150,8, rendered by Pruden 1988: 437 as "thirty-six points of the

Master".58T 1558 at T XXIX 54b12, cf. also Hirakawa 1978: 97, who gives the Tibetan equivalent as ston

pa'i bka', the "teacher's pronouncement(s)".The Mahvibh in T 1545 at T XXVII 718a25also speaks of .

46


15/24

translation speaks of ,59which suggests ~ ntapadni, "peaceful

positions", or perhaps "paths of tranquillity".

The next example is from the 171stdiscourse of theMadhyama-gama, the ,which begins a treatment of karma with an announcement by the Buddha,in which he told nanda that if he (and the other monks) would hear this greatexposition on karma, they would develop increasing mental tranquillity andhappiness in regard to the Tathgata.

9) 60

In the Pli counterpart, this sentence does not seem to have been preserved in

full, as the passage just reads "if you, nanda, would hear the detailed greatexposition on karma by the Tathgata", 'sace tumhe, nanda, sueyythatathgatassa mahkammavibhaga vibhajantass'ti, at which point the sentenceends.61The way this sentence reads one has the impression that it could have beentruncated, since even though it begins with the hypothetical indeclinable "if", sace,it ends without explaining what will happen "if" the monks and nanda hear thedetailed great exposition on karma.

Here the Madyhama-gama version offers a more complete version of theBuddha's statement, and with its help the Pli sentence could be restored to some-thing like: sace tumhe, nanda, sueyytha tathgatassa mahkammavibhagavibhajantassa, tatra vo, nanda, tathgate citta bhiyyosomattya pasdeyya

pmojja labheyya.

Another example where the Chinese translation is of help to clarify a Plipassage can be found in the 174thdiscourse of theMadhyama-gama, the.This discourse illustrates the danger of indulgence in sensual pleasures with theexample of a slatree that is gradually overgrown by a creeper. The Pli versionof this simile describes that when at first the seed of this creeper falls down in the

vicinity of the slatree, the worries of the deva that lives in the tree will be ap-peased by his friends who tell the deva that the seed will probably be eaten bysome animal, or carried away, or else the seed of such a creeper might become a

59T 1559 at T XXIX 211b13.60M 171 at T I 707a19: (adopting the , , and variant reading for )61MN 136 at MN III 209,12. The same pattern recurs again in regard to a similar proclamation

made by the Buddha according to AN 6:62 at AN III 404,5: 'sace tumhe, nanda, sueyythatathgatassa purisindriyani vibhajantass'ti, where again the Madhyama-gamacounterpart in M 112 at T I 601a15completes the sentence in a way closely resembling M171: a reading which in this case is,moreover, found similarly in another parallel, the individual translation T 58 at T I 854b 7:

47


16/24

"no-seed", abja v pan' assa.62 The last of these possibilities is not entirely

clear, since one might wonder what the implications could be of a seed becominga "no-seed". Here the Madhyama-gamaparallel helps to clarify the implicationof this imagery, as it describes how this seed might "rot and not be a seed [anymore]".

10)63

The next example stems from the 190thdiscourse of theMadhyama-gama, the, which treats a series of meditative experiences that lead to the realizationof emptiness in its ultimate and supreme sense. In the context of this series ofmeditative experiences, this discourse speaks of an "unconscious concentration of

the mind", but then continues to describe a form of practice that is based on"knowing" the nature of this attainment.

11) 64

Its Pli and Tibetan parallels, however, speak instead of a "signlessconcentration of the mind", animitta cetosamdhi,65or of the signless element,mtshan ma med padbyings,66a reading that would fit the context much better.

The appearance of a reference to "unconscious", , in a context where thismeaning does not fit too well seems to be a recurrent issue, since several similarcases can be found. One such case is the , which refers to an"unconscious concentration", 67 while its Pli parallel speaks of "signlessconcentration" instead.68In this case, the Pli version's reading is supported by aquotation from the same discourse in the Karmasiddhi-prakaraa, which indeedspeaks of "signless" concentration.69

Quite a number of similar occurrences can be found in other Chinese discoursetranslations,70 which indicate that the two characters and were prone to

62MN 45 at MN I 306,12.63M 174 at T I 711c11.64M 190 at T I 737c3.65

MN 121 at MN III 107,29

.66Skilling 1994: 172,5.67M 211 at T I 792b12: . Notably, subsequent occurrences of this expression in M 211

have a variant reading as .68MN 43 at MN I 296,32.69T 1609 at T XXXI 784b18: 70Choong 1999: 116 note 220 draws attention to another instance found in S 272 at T II 72a 26,

where a counterpart to animitta in SN 22:80 at SN III 93,23reads , "signless", but has avariant reading as , "unconscious". The reverse case occurs in M 146 at T I 657c4as partof a description of sense-restraint, which in the Pli version MN 27 at MN I 180,27speaks of the"sign", nimitta, but in the Chinese reads , "perception", with the variant reading , "sign". A

48


17/24

being confused with each other, so that the correct reading needs to be established

in each case based on the context. A confusion of and could easily happenduring translation, as the two characters are not only fairly similar in writing,differing only on the presence or absence of the heart radical, but also had a rathersimilar pronunciation in Early Middle Chinese, as they do still today.71Due to therelated meaning of the two terms, such an error would then easily escape beingnoticed during a later checking of the translation. Thus, this finding corroboratesthe suggestion made by Master Yn-shn (1986: 61), who based on his extensivereadings in the Chinese Tripiaka, comes to the conclusion that the "unconsciousconcentration of the mind" should simply be treated as an alternative rendering forthe signless concentration of the mind.72

An instance where the Madhyama-gama version helps to better understand aPli passage can be found in the 192nddiscourse, the . The Chineseand Pli versions of this discourse describe a situation where a monk goesbegging during a stormy night. During a flash of lightning, a woman suddenlysees this monk searching for alms and is thoroughly terrified, believing him to be

complementary case is M 187 at T I 733c19+22, which uses the character , "sign", to describethe practice of sense-restraint, but then notes , "perception", as a variant reading for the samecontext. Again in M 169 at T I 701c1the expression "not connected with benefit", ,has as a variant for , on adopting which the expression as a rendering of ~ sa + yuj,

or sa + bandh would lose its sense. M 34 at T I 475b8+16refers to the absence of pride withthe expression , while the same discourse M 34 at T I 475b2 refers to thesame absence of pride with the expression , yet another instance where thecharacters and appear to have been confused with each other. Another instance is T 92 atT I 916c8, where the character , "sign", occurs in a description of overcoming all perceptionsof form in order to reach the immaterial attainments, with the better fitting , "perception", as avariant reading. Again, E 24.8 at T II 629b1refers to the fourth immaterial attainment as , with the better fitting , as a variant reading, a reading confirmed in E 24.8 at TII 629b3+22+24. Another example occurs in a description of a meditation practice undertakenregularly by the Buddha in T 76 at T I 884b17, according to which he practised "unconscious"concentration,

, with the better fitting "signless concentration", , as a variantreading. Yet another example can be found in T 6 at T I 180a16, which speaks of the samdhi

used by the Buddha to overcome an illness as , "concentration of not giving at-tention to numerous perceptions", while the corresponding Sanskrit fragment S 360 folio 171V4 in Waldschmidt 1950: 18 and the Tibetan version in Waldschmidt 1951: 195,1speak insteadof "not giving attention to any signs", sarvanimi(ttnm amanasikrd)and mtshan ma thamscad yid la mi bya bar. The idea to not give attention to "signs", sabbanimittnam amanasikr,occurs also in the corresponding Pli passage in DN 16 at DN II 100,16, though not explicitly asthe means used by the Buddha to overcome his illness. The (Mla-)sarvstivda Vinayasimilarly speaks of the "signless concentration" at this point, T 1451 at T XXIV 387a22: .

71Cf. Pulleyblank 1991: 337 and 338 or Unger 1989: 89.72.

49


18/24

an evil spirit. According to the Pli, the frightened woman makes a rather cryptic

remark, exclaiming that the mother and father of this monk have died, bhikkhussatu mri, bhikkhussa mtu mri.73As it stands in the Pli version, this remark isnot easily intelligible.

According to the commentary, the intended meaning is that if the monk's parentswere still alive, he would not need to go in search for food during the night.74Thiscommentarial explanation does not seem to fit the situation too well, as it was ageneral custom for monks and recluses in ancient India to subsist on begging foodas part of their way of life. Thus, the idea that they should be fed by their parentsinstead of begging their food would not be the type of reasoning to be expected ofa woman in ancient India.

The solution to this cryptic passage can be found in the Madhyama-gama

version, where the reference to the monk's parents comes as part of a set of cursesspoken by the frightened woman, in which she expressed her anger by wishingthat his mother and father may pass away, and that his whole clan may meet withdestruction

12)75

In this way the Madhyama-gama discourse clarifies that the reference to thedeath of the monk's parents was part of a curse spoken by the frightened woman, apresentation that seems to fit the context better than the explanation offered in thePli commentary.

Continuing with another example related to the conduct of monks, the 194thdis-course in the Madhyama-gama, the , describes the behaviour of amonk who is investigated for some misdeed. According to its description, onetype of monk displays anger and then says that he wants to act in accordance withthe wishes of the sagha, while another type of monk does not display anger anddoes not say that he wants to act in accordance with the wishes of the sagha.

13) 76 77

The Pli version has these in a different manner, as according to its presentation

73MN 66 at MN I 449,1: (Be-MN II 112 reads mr). Trenckner 1993: 567 comments that "the textno doubt purports to make the woman speak a sort of patois". amoli 2005: 552 renders thispassage as "a bhikkhu whose ma's died and whose pa's died", with Bodhi in ibid. p. 1270 note672 explaining that "the utterance ... appears to be a very colloquial Pali".

74Ps III 165.75M 192 at T I 741b16.76M 194 at T I 748b28.77M 194 at T I 748c18.

50


19/24

the monk who displays anger is the one who is also unwilling to act in accordance

with the wishes of the sagha, and the one who does not display anger is the onewho says that he is willing to act in accordance with the wishes of the sagha.78This is in fact what one would expect, so that the present instance could beanother instance, similar to the first example given above, where amisunderstanding of a sandhimay have led to a confusion between the positiveand the negative versions of a statement.

My next example comes from the 211stdiscourse in theMadhyama-gama, the, which explains that when someone enters the attainment of cessation,first the bodily formations cease, then the verbal formation, and then the mentalformation.To understand this proposition it needs to be kept in mind that in the

early discourses the bodily formation stands for breathing in and out, the verbalformation for initial and sustained mental application, and the mental formationfor perception and feeling.

14)79

According to the parallel, the Cavedalla-sutta of the Majjhima-nikya,however, first the verbal formation ceases (initial and sustained mentalapplication), followed by the bodily formation (breathing in and out), and finallythe mental formation ceases (perception and feeling).80

In relation to the emergence from cessation, the same difference recurs, though

obviously in the reverse order. According to the Pli version the verbal formationis the last to arise, while according to the Chinese presentation the bodilyformation arises last.

In relation to the sequence in which the three formations ceases, it seems thatwhen proceeding through the jhnas in order to attain cessation, the verbalformation of initial and sustained mental application will be left behind on attain-ing the second jhna, while the bodily formation of in- and out-breathing willonly cease with the attainment of the fourthjhna.81This would correspond to thesequence proposed in the Pli version of the Cavedalla-sutta. This sequencereceives further support from a discourse in the Sayukta-gama and its Pliparallel. According to these two discourses, on attaining cessation the first forma-

78MN 65 at MN I 442,31and MN I 443,10.79M 211 at T I 792a8, on the different roles taken by its protagonists in the Chinese and Pli

versions cf. Anlayo 2007.80MN 44 at MN I 302,4. For an examination of the difficulties involved in explaining emergence

from the attainment of cessation and how different Buddhist schools attempted to tackle thisproblem cf. Griffith 1991.

81DN 33 at DN III 270,18and AN 10:20 at AN V 31,25speak of the tranquillisation of the bodilyformation with the fourth jhna, at which point according to SN 36:11 at SN IV 217,8 allbreathing ceases.

51


20/24

tion to cease is indeed the verbal formation of initial and sustained mental

application.82

Besides, the Mahvibh also agrees with the sequence proposedin the Pli version of the Cavedalla-sutta.83Thus this part of the Madhyama-gama discourse appears to offer the less probable reading. It could easily beimagined how the present reading in the Madhyama-gama version could havecome into being due to an error during oral transmission. The standard triad body-speech-mind, which is such a recurrent feature in the early discourses, couldeasily have influenced the memory of the reciters and caused them to recite thepassage in the standard sequence to which they were used to, instead of adoptingthe unusual sequence verbal-bodily-mental, even though in the present contextthis sequence would be the correct presentation.

The 213th

discourse in the Madhyama-gama, the ,describes that theBuddhist monks were independent and free from wishes and would life the holylife for as long as their body would last, a description in which the same discoursealso explains that the monks would "protect food from others' wives, like deer".

15)84

As it stands, this description is difficult to understand. According to the Pliversion, however, the monks are "dependent on others with a mind like deer",paradavutta migabhtena cetas.85This suggests that theMadhyama-gamaver-sion's reference to "others' wives" may be due to a mistaking of parada for

paradra, "the wife of another", while the idea of "protection" could be due tomistaking vutta for vta, "stopped", "checked", "held back", or else could be ascribe's error for , "to get". Even though the translator(s) appear(s) to havemisunderstood the expression paradavutta, perhaps through knowledge of somecommentarial explanation he/they knew that the idea of "depending on others"was involved, an idea he?they may then have attempted to bring out with theimagery of "food from others' wives". The difficulties of the translator(s) areunderstandable, as the expression paradavutta is difficult and while thecommentaries explain it to mean "dependent on others",86 the PTS dictionaryrenders it literally as "fond of being prepared" and then explains it to mean "apt,

active, alert".

87

A version of this passage in theKudrakavastu of the (Mla-)Sar

82S 568 at T II 150b20and SN 41:6 at SN IV 294,8.83T 1545 at T XXVII 780c25.84M 213 at T I 796a29, with avariant readingfor.85MN 89 at MN II 121,21.86 Ps III 166: 'paradavutt'ti parehi dinnavuttino.87

Rhys Davids 1993: 420 s.v. parada;on the difficulties of this expressioncf. also Horner 1975:259 note 2.

52


21/24

vstivdaVinaya is also far from clear, as it reports that the monks "always feel

fear, like deer in the forest".88

As a last example in my presentation I would like to take up the 214thdiscoursein the Madhyama-gama, the . This discourse describes how a kinginquired from nanda if the Buddha would undertake any bodily deed that couldbe censured by other Brahmins and recluses. In reply, nanda explained that theBuddha would not undertake a bodily deed that could be censured by "wise"Brahmins and recluses.

16) 89

This specification is quite significant, since foolish people can blame even asaint. Thus, the point of this additional qualification seems to be that nandawanted to distinguish between the righteous censure by "wise" recluse andBrahmins and unjustified criticism that was sometimes raised by contemporaryrecluses and Brahmins against the Buddha. The appropriateness of thisqualification can be seen in the Chinese and Pli versions of the present discourse,as both record that the king expressed his appreciation for nanda's reply, in factaccording to the Pli version he proclaimed that what he had not been able to ac-complish with his question, nanda had accomplished with his reply, ya himaya ... nsakkhimha pahena paripretu, ta ... yasmat nandena

pahassa veyykaraena pariprita.90

On reading the PTS edition of this discourse, however, this sentence is a littlepuzzling, since according to its presentation the king already used thequalification "wise" in his question. A closer inspection of the different Plieditions brings to light that though the Siamese edition agrees with the PTSedition,91the Burmese and the Ceylonese editions differ, in that according to themthe king had not yet used the qualification "wise" in his question. 92Thus, in thiscase the Madhyama-gama version helps to show which of the Pli editionscarries the preferable reading.

By way of concluding my comparative notes on theMadhyama-gama, I wouldlike to suggest that these few selected examples show how much our study of thediscourses can gain i f the Chinese and Indic versions are placed side by side. Thus,

just as according to de Jong (1968: 15) "no student of Buddhism, even if he isinterested only in Indian Buddhism, can neglect the enormous corpus of Chinese

88T 1451 at T XXIV 237c29:.89M 214 at T I 798a13.90M 88 at M II 114,7.91The PTS edition at M II 113,33and Se-M II 500 readsamaehi brhmanehi vihi.92Be-M II 315 and Ce-M II 542 only read samaehi brhmanehi.

53


22/24

translations", I would like to propose that "no student of Chinese Buddhism can

neglect the study of the Indic parallel versions", in order to minimize the risk ofcoming to conclusions that are based on transmission or translation errors.

Abbreviations:(In the case of Chinese and Pli sources, quotations are according to the Taish and PTS editionsby giving first the discourse by number and then its location by volume, page and line; in the caseof Tibetan sources, quotations are to the location in the Derge and/or Peking editions).AN Aguttara-nikyaBe Burmese editionCe Ceylonese editionD Dirgha-gama(at T 1)

D2 Dirgha-gama(preserved in Sanskrit fragments)D Derge editionDN Dgha-nikyaE Ekottarika-gama (at T 125)M Madhyama-gama(at T 26)MN Majjhima-nikyaPs PapacasdanQ Peking editionS Sayukta-gama(at T 99)S2 'other' Sayukta-gama(at T 100)SHT III see Waldschmidt 1971SHT IX see Bechert 2004Se Siamese edition

SN Sayutta-nikyaSN2 Sagthavagga of the Sayutta-nikya, new PTS edition by Somaratne (1998)T TaishVibh VibhagaVin Vinaya

References:Akanuma, Chizen 1990 (1929): The Comparative Catalogue of Chinese gamas & Pli Nikyas,

Delhi: Sri Satguru Publications.Anlayo 2005: "Some Pli Discourses in the Light of Their Chinese Parallels", inBuddhist Studies

Review, vol. 22 no.1 pp. 1-14, and vol. 22 no. 2 pp. 93-105.Anlayo 2006: "The Ekottarika-gama Parallel to the Saccavibhaga-sutta", in Buddhist Studies

Review, vol. 23 no. 2.Anlayo 2007: "Who said it? Authorship Disagreements between Pli and Chinese Discourses", in

Indica et Tibetica 65, Festschrift fr Michael Hahn zum 65. Geburtstag von Freundenund Schlern berreicht, J. U. Hartmann et al. (ed.), Wien, pp. 1-14.

Bagchi, Prabodh Chandra 1927:Le Canon Bouddhique en Chine, Paris: Geuthner, vol. 1.Bapat, P.V. 1969: "Chinese Madhyamgama and the Language of its Basic Text", in Dr. Satkari

Mookerji Felicitation Volume, B.P. Sinha (ed.), Varanasi: Chowkhamba Publications, pp.1-6.

Bechert, Heinz 1991: "Methodological considerations concerning the language of the earliestBuddhist tradition", inBuddhist Studies Reviewvol. 8 pp. 3-19.

Bechert, Heinz (ed.) 1995: When did the Buddha live?, Delhi: Sri Satguru.

54


23/24

Bechert, Heinz (ed.) 2004: Sanskrithandschriften aus den Turfanfunden. Teil 9: Die

Katalognummern 2000-3199(beschrieben von Klaus Wille), Stuttgart: Franz Steiner.Choong, Mun-keat 1999: The Notion of Emptiness in Early Buddhism, Delhi: Motilal Banarsidass.de Jong, J.W. 1968: Buddha's Word in China, 28th George Ernest Morrison Lecture, Canberra:

Australian National University.Enomoto, Fumio 1984: "The Formation and Development of the Sarvstivda Scriptures", in

Proceedings of the Thirty-First International Congress of Human Sciences in Asia andNorth Africa, T. Yamamoto (ed.), Tokyo: Th Gakkai, pp. 197-198.

Enomoto, Fumio 1986: "On the Formation of the Original Texts of the Chinese gamas", inBuddhist Studies Review, vol. 3 pp. 19-30.

Glass, Andrew 2006: Connected Discourses in Gandhra: A Study, Edition, and Translation ofFour Sayuktgama-Type Stras from the Senior Collection, PhD dissertation,University of Washington, published at http://www.andrewglass.org/phd.php (accessedon 04-08-2006).

Gnoli, Raniero 1978: The Gilgit Manuscript of the Saghabhedavastu, vol. 2, (Serie OrientaleRoma, vol. 49), Rome: Istituto Italiano per il Medio ed Estremo Oriente.Griffith, Paul J. 1991 (1986): On Being Mindless: Buddhist Meditation and the Mind-Body

Problem, Illinois, La Salle: Open Court.Hartmann, Jens-Uwe 2000: "Zu einer Neuen Handschrift des Drgh gama," in

Vividharatnakarandaka: Festgabe fr Adelheid Mette, Chojnacki et al (ed.), Swisttal-Odendorf, pp. 359-367.

Hartmann, Jens-Uwe 2004: "Contents and Structure of the Drghgama of the (Mla-)Sar-vstivdins" in Annual Report of the International Research Institute for AdvancedBuddhology, Tokyo: Soka University, vol. 7 pp. 119-137.

Hirakawa, Akira 1978: Index to the Abhidharmakoabhya, Tokyo: Daizo ShuppanKabushikikaisha, vol. 3.

Hirakawa, Akira 1997:Buddhist Chinese-Sanskrit Dictionary, Tokyo: Reiyukai.

Horner, I.B. 1967: The Collection of the Middle Length Sayings, London: Pali Text Society, vol. 1.Horner, I.B. 1975: The Book of the Discipline, London: Pali Text Society, vol. 5.Horner, I.B. 1990 (1930): Women under Primitive Buddhism, Delhi: Motilal Banarsidass.Hultzsch, E. 1925:Inscriptions of Asoka, Oxford: Clarendon Press.Karashima, Seishi 1992: The Textual Study of the Chinese Versions of the Saddharmapua-

rkastra, Tokyo: Sankibo Press.Kudo, Noriyuki 2004: The Karmavibhaga, Tokyo: Soka University.Lamotte, tienne 1967: "Un Stra Composite del'Ekottargama", in Bulletin of the School of

Oriental and AfricanStudies, Oxford, vol. 30 pp. 105-116.Lancaster, Lewis R. 1979: "Buddhist Literature: Its Canons, Scribes, and Editors" in The Critical

Study of Sacred Texts, W.D. O'Flaherty (ed.), Berkeley, pp. 215-229.Lvi, Sylvain 1896: "Notes sur des Inscriptions de Piyadassi - Le Lghulovda de l'Edit de

Bhabra", inJournal Asiatique ser. 9 vol. 8 pp. 475-485.

L, Cheng 1963: "gama", inEncyclopaedia of Buddhism, Ceylon, vol. 1 pp. 241-244.Mayeda, Egaku 1985: "Japanese Studies on the Schools of the Chinese gamas", in ZurSchulzugehrigkeit von Werken der Hnayna-Literatur, H. Bechert (ed.), Gttingen:Vandenhoeck & Ruprecht, vol. 1 pp. 94-103.

Mette, Adelheid 1995: "The Synchronism of the Buddha and the Jina Mahvra and the Problem ofChronology in Early Jainism", in When did the Buddha Live?, H. Bechert (ed.), Delhi: SriSatguru, pp. 179-183.

Minh Chau, Thich 1991: The Chinese Madhyama gama and the Pli Majjhima Nikya , Delhi:Motilal Banarsidass.

Mittal, Kusum 1957: Dogmatische Begriffsreihen im lteren Buddhismus I; Fragmente desDaottarastra; (Sanskrittexte aus den Turfanfunden), Berlin: Akademie Verlag, vol. 4.

55


24/24

amoli, Bhikkhu 2005 (1995): The Middle Length Discourses of the Buddha, Boston: Wisdom.

Norman, K.R. 1989: "The Pli language and scriptures", in The Buddhist Heritage, Skorupski (ed.),pp. 29-53.Norman, K. R. 1992: "The Value of the Pli Tradition", in Collected Papers, Norman, Oxford:

PTS, vol. 3 pp. 33-44.Pradhan, P. 1967: Abhidharmakoabhya, (Tibetan Sanskrit Works Series 8), Patna: K.P.

Jayaswal Research Institute.Prasad, Chandra Shekhar 1985: "Some Reflections on the Relation between the gamas and the

Nikyas", in Proceedings and Papers of the Second Conference of the InternationalAssociation of Buddhist Studies, Nalanda pp. 131-140.

Pruden, Leo M. 1988b:Abhidharmakoabhyam, Berkeley: Asian Humanity Press, vol. 2.Pulleyblank, Edwin G. 1991: Lexicon of Reconstructed Pronunciation in Early Middle Chinese,

Late Middle Chinese and Early Mandarin, Vancouver: UBC Press.Rhys Davids, T.W. 1993 (1921-25):Pli-English Dictionary, Delhi: Motilal Banarsidass.

Shukla, Karunesha 1973: rvakabhmi of crya Asaga, Patna: Jayaswal Research Institute.Skilling, Peter 1994:Mahstras: Great Discourses of the Buddha, Oxford: PTS, vol. 1.Trenckner, V. 1993 (1888): The Majjhima Nikya, Oxford: PTS.Unger, Ulrich 1989: Glossar des Klassischen Chinesisch, Wiesbaden: Harrasowitz.von Hinber, Oskar 1982: "Upli's Verses in the Majjhimanikya and the Madhyamgama", inIn-

dological and Buddhist Studies, Canberra, pp. 243-251.von Hinber, Oskar 1993: "From Colloquial to Standard Language. The Oral Phase in the

Development of Pli", inPremier Colloque tienne Lamotte, Louvain-la-Neuve, pp. 101-113.

Wagle, Narenda 1985: "The Gods in Early Buddhism in Relation to Human Society" inNew Pathsin Buddhist Research, A.K. Warder (ed.), Durham, pp. 57-80.

Waldschmidt, Ernst 1950/vol. 1, 1951/vol. 2: Das Mahparinirvastra, Berlin: AkademieVerlag.

Waldschmidt, Ernst 1956: Das Mahvadnasutra, ein Kanonischer Text ber die sieben letztenBuddhas,Berlin: Akademie Verlag, vol. 2.Waldschmidt, Ernst (ed.) 1971: Sanskrithandschriften aus den Turfanfunden. Teil 3:

Katalognummern 802-1014 (unter Mitarbeit von Walter Clawitter und Lore Sander-Holzmann), Stuttgart: Franz Steiner.

Waldschmidt, Ernst 1980: "Central Asian Stra Fragments and their Relation to the Chinese ga-mas", in H. Bechert (ed.) The Language of the Earliest Buddhist Tradition, Gttingen:Vandenhoeck & Ruprecht, pp. 136-174.

Yn-shn 1983 (1962): [The Compilation of the Early Buddhist Canon],Taipei: .

Yn-shn 1986 (1985):[Exploration of Emptiness], Taipei:.

56

comparative notes on the madhyama-āgama

Documents